Kinds, essences, powers

What is the new essentialist asking us to accept? Not that there are natural kinds, nor that there are intrinsic causal powers. These things could be accepted without a commitment to essentialism. They are asking us to accept something akin to the Kripke-Putnam position: a metaphysical theory about kind-membership in virtue of essential properties. But Salmon has shown that there is no valid argument for the Kripke-Putnam position: no valid inference that gets us from reference to essence. Why then should we accept essentialism? A remaining reason is Ellis's argument by display: we should buy essentialism because of the benefits it will bring. But are these benefits real? The problem is that the putative benefits of essentialism - that the laws of nature are necessary, that the problem of induction is solved, and so on - look actually to be the assumptions of Ellis's theory. If that is the case, there is no real benefit to be gained from adopting the theory. The argument for essentialism is therefore underdetermined and it remains possible to accept natural kinds into one's ontology without accepting their corresponding essences.